Zwischen Definition und Empirie

Theoretische Probleme der Fragebogenkonstruktion

In: Vierteljahrsschrift für wissenschaftliche Pädagogik
Matthias Borgstede Otto-Friedrich-Universität Bamberg, Institut für Erziehungswissenschaft, Allgemeine Pädagogik Markusplatz 3, D-46045 Bamberg

Search for other papers by Matthias Borgstede in
Current site
Google Scholar
Download Citation Get Permissions

Access options

Get access to the full article by using one of the access options below.

Institutional Login

Log in with Open Athens, Shibboleth, or your institutional credentials

Login via Institution


Buy instant access (PDF download and unlimited online access):



Between Definition and Empirical Research. Theoretical Problems of Questionnaire Design

Within the field of quantitative educational science, theoretical terms are often defined operationally by scales or questionnaires, which have been constructed using the formal background of Classical Test Theory. This article takes a critical perspective on the current modus operandi of scale construction.

It is argued that classical quality criteria (reliability, validity, and objectivity) are usually represented by empirical correlations between scales or items of a scale. Apart from some theoretical inconsistencies, using these criteria (e.g. Chronbach’s Alpha) to optimize a scale results in a set of items with high intercorrelations, regardless of their content or meaning.

Because high inter-item correlations are neither necessary nor sufficient for an adequate operational definition, the utility of the classical approach to scale construction is questioned on a fundamental level. It is further argued that operational definitions do not pose a statistical problem in the first place, but a theoretical problem. Hence, we do not need a method to optimize scales for high inter-item-correlations, but rather a way to translate theoretical terms (which may be inherently vague) to empirically observable phenomena. This is accomplished by the semantic approach to psychological tests by Buntins, Buntins and Eggert (2016), which provides a formal alternative to classical quality criteria.

Content Metrics

All Time Past Year Past 30 Days
Abstract Views 811 172 9
Full Text Views 199 44 4
PDF Views & Downloads 225 71 2